next up previous contents
Next: 7 Creating a Spun-up Up: UsersGuide Previous: 5 CLM3.0 Data Structures   Contents


6 CLM3.0 Surface Dataset Formats

As mentioned in section 4.2, CLM3.0 now supports two surface-data formats. The new format differs from the old format in only two variables:

The new format surface dataset will be created if the following namelist variable is set:

    mksrf_all_pfts = .true.

The original format surface dataset will be created by default, or if

    mksrf_all_pfts = .false.

The surface data in its new format provides more flexibility than in the original format by allowing the user to decide the number of dominant pfts per gridcell for their simulation without creating a new surface-data file each time this number changes. This is possible because all pfts found in the raw data are included in the new format surface dataset in the order that they are listed in the pft-physiology file FPFTCON. Using the new format surface-data file as input, the model selects at run-time the 4 dominant natural pfts and places them on a natural vegetation landunit, while it places crops separately on a crop landunit.

The user may change at compile time the model parameter that specifies the number of dominant pfts used in the model. The parameter, maxpatch_pft in module clm_varpar.F90 in the src/main directory determines the maximum number of vegetated pfts in the naturally vegetated landunit. Currently this is set to 4. If the user were to increase this value to 6 and recompile the code, then the model would select the 6 dominant natural pfts and place them on the natural vegetation landunit. In the original format, after recompiling the code, a new surface dataset would need to be created, which would contain the 6 dominant pfts (both natural and crop).

Similar to maxpatch_pft, the parameter maxpatch_cft specifies the maximum number of crop pfts in the crop landunit in the new format. Currently this number is set to 2 and corresponds to corn and wheat vegetation types. In CLM3.0, these vegetation types for corn and wheat are currently hard-wired to values of 15 and 16, respectively, which are identical in terms of their physiological properties. This hard-wiring of values will be removed in future releases.

The new format surface-data file will facilitate changing the land cover in the middle of a simulation. It is important to note, however, that such a capability will require additional code development which is not currently in the CLM3.0 release. The presence of distinct natural and crop landunits will also allow the separate treatment of these landunits using different modules. For example, the dynamic vegetation model (DGVM) in CLM3.0 may be used for the natural landunit, while a crop model could be used for the crop landunit. Although no crop model is included yet and only natural vegetation is permitted when DGVM is active in the CLM3.0 release, the new surface dataset form will facilitate the introduction of these changes in the code.

The new surface-data format and the separate vegetation landunits lead to small changes in CLM's simulated fluxes to the atmosphere. As a result, one will see small changes in the simulated climate in CAM or CCSM mode. Users should decide to work with one of the surface-data formats and not switch between such formats in the middle of a study. However, if a user desires to do surface dataset development, they are strongly encouraged to use the new surface dataset format.


next up previous contents
Next: 7 Creating a Spun-up Up: UsersGuide Previous: 5 CLM3.0 Data Structures   Contents
csm@ucar.edu